With the development of computer network technology, people can access millions of web sites and billions of web pages over the internet. Like newspapers, these various web pages have layouts. FIG. 1 shows a typical layout of web page 10, including such layout elements as news header 11, news body 12, related links 13, top banner 14, vertical banner 15 and bottom banner 16 as well as possible rolling advertisements (not shown). Web pages developed and deployed by different designers and developers may have different template, and various parts of the web layout/elements often indicate different information, e.g., the navigator, the content, and the advertisement and so on.
For web pages, the elements in the layout may have different values for different audience. For example, an ordinary user usually only cares about news content on the web page, whereas an advertiser may care more about the ad information. However, currently when the web layout/elements are downloaded to the user's browsers, they are not prioritized according to their importance and business values. Rather, their download and display are fully dependent upon the download speed of the web layout/elements, and they are downloaded to the browser purely randomly, even during the network peak time. Due to the present multi-thread parallel download mechanism during web page downloading, the web elements are non-discriminatively served, which means it might take a long time to download contents that turn out to be of no interest to the user, and it might be not possible for the contents that the user is interested in to be finally downloaded and displayed (this largely depends on the status of the application server, network, client and browser). Likewise, the advertisers have similar demand, although their prioritization of the elements might be totally different from that of ordinary users.
Presently, there are only guidelines, or principles to help the design of the layout of the web pages/web elements, i.e., the fonts, the images, etc., but no system provided for developers to annotate the different parts of the layout with different priority level and priority type. There is also no system to enforce the prioritization of web content during runtime. In addition, in high volume condition, there is optimization mechanism according to the different service level agreement of customers (for example, different user have different speed for downloading and display as well as different update cycle), however, this does not relate to the differentiation of the web contents itself. Moreover, the current optimization granularity is cross page level and highly dependent on the web application server.
Therefore, it is desirable to implement systematic optimization consideration of the collaboration of both client and application server side.